Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

True aperture? A Quick Way to Measure!

This topic has been archived. This means that you cannot reply to this topic.
171 replies to this topic

#51 StarStuff1

StarStuff1

    Skylab

  • *****
  • Posts: 4,155
  • Joined: 01 Apr 2007

Posted 22 September 2009 - 11:59 PM

Hi Ken,

The Swifts I have are the older No 804 model with 445 ft @ 1000 yds printed on them. This works out to about 8.5°. I have read that these were designed primarily for serious birders. The stars in my 804 start softening about 2/3 from the center to the edge and get worse. Still, the edges are better than many binos I have seen with smaller fovs. When you hold the Swifts up to the stars and take in the view by mostly looking at the center of the fov you don't notice the soft edge.

The bino the second link goes to is for a newer version with lighter weight and only a very slightly smaller fov. Being waterproof and rubber armored doesn't hurt, either. I can't really comment on the quality of view in this one since I have never looked through the 820 but I suspect the view would be very satifactory. The reviews I have read indicate that this is a very good bino for the money.

#52 Tony Flanders

Tony Flanders

    Hubble

  • *****
  • Posts: 18,776
  • Joined: 18 May 2006

Posted 23 September 2009 - 07:25 AM

Parallax error during measurement will not be an issue when the circle of light is projected *onto* a ruler.


Quite right -- I can't believe I didn't think of that (doh).

Since I do a lot of work on star charts and diagrams, I have a huge collection of rulers and other drafting tools -- all of them made out of clear plastic, which is *not* the right tool for this job. Using an old-fashioned wooden ruler, projecting the circle directly onto the ruler instead of onto a white surface behind the ruler, does indeed eliminate parallax.

I'm now sure that I can measure the aperture +-1 mm, which is ample accuracy in practice. Probably +-0.5 mm, though I'm not so sure about that.

It would be nice to have really high accuracy, because this exact same method can be used in reverse to measure the exit pupil. And given aperture and exit pupil, you can then measure magnification. (Average magnification over the entire FOV, anyway, which is *not* the same as magnification in the center of the field.)

Unfortunately, an 0.5-mm error measuring the exit pupil would give a 10% error measuring the magnification of binoculars with a 5-mm exit pupil. To be really useful, the exit pupil needs to be measured +-0.1 mm.

Although not nearly as accurate as measuring the aperture for an on-axis beam, this method also gives an excellent feel for how the beam is vignetted off-axis. Move the flashlight laterally, away from the central axis of the objective, then re-aim the beam at the center of the objective, so that it's illuminating the whole EP more or less evenly, but from an angle. You will see that the projected beam -- the effective aperture -- is no longer circular, once you get far enough off-axis. Other circles, and possibly the square prism edges, will start to clip the beam.

This isn't 100% accurate, for a variety of reasons. But it does hint at some purely geometric ways to measure the percentage illumination of the FOV once you get away from the center of the field -- the subject of an earlier note string.

Geometric methods can't take the place of measuring light throughput with a meter, because they ignore issues of light scatter from refractive surfaces, imperfect reflection from prisms, and absorption in the glass. However, geometric methods are *much* easier to apply than photometry!

More on this once I've done some experiments.

#53 Andresin150

Andresin150

    Surveyor 1

  • *****
  • Posts: 1,511
  • Joined: 14 Aug 2007

Posted 23 September 2009 - 09:32 AM

Not that I expected something different, but my giant Fujis measured 150mm exactly.... very easy to measure since you can project the beam in a paper placed over the dew shields, so it its perfectly aligned.
I'll post other "results" later

#54 EdZ

EdZ

    Professor EdZ

  • *****
  • Posts: 18,849
  • Joined: 15 Feb 2002

Posted 23 September 2009 - 11:25 AM

It would be nice to have really high accuracy, because this exact same method can be used in reverse to measure the exit pupil. And given aperture and exit pupil, you can then measure magnification. (Average magnification over the entire FOV, anyway, which is *not* the same as magnification in the center of the field.)

Unfortunately, an 0.5-mm error measuring the exit pupil would give a 10% error measuring the magnification of binoculars with a 5-mm exit pupil. To be really useful, the exit pupil needs to be measured +-0.1 mm.


Well, basically that is the method I've been using for years to map exit pupil. The laser test.

This would only allow you to measure exit pupil IF you took the measurement at precisely the eye relief distance, which adds another measure of difficulty. As you move even 6" to 12" away, the circular projection grows to several inches large. Frankly, I wouldn't use this method for measuring exit pupil.

edz

#55 Tony Flanders

Tony Flanders

    Hubble

  • *****
  • Posts: 18,776
  • Joined: 18 May 2006

Posted 23 September 2009 - 01:09 PM

This would only allow you to measure exit pupil IF you took the measurement at precisely the eye relief distance, which adds another measure of difficulty. As you move even 6" to 12" away, the circular projection grows to several inches large.


I don't see that. If you focus the device for infinity and throw a collimated beam into the objective, you should get a collimated beam out the eyepiece.

The key is that the light source has to be effectively at infinity in order to produce a collimated beam. But as with Glenn's test, you can get pretty close to a collimated beam simply by placing the light source far away -- around 10X the focal length of the lens you're illuminating.

An alternative, as Glenn suggests, is to use a telescope with aperture larger than the binocular to produce the collimated beam. In this case, you put the light source at the telescope's focal plane and the collimated beam shoots out the telescope's aperture.

Frankly, I wouldn't use this method for measuring exit pupil.


Me neither, because I don't have the technology to measure the beam accurate to 0.1 mm.

#56 GlennLeDrew

GlennLeDrew

    Hubble

  • *****
  • topic starter
  • Posts: 15,850
  • Joined: 17 Jun 2008

Posted 23 September 2009 - 06:23 PM

Following the slightly new direction of exit pupil measurement...

I measure exit pupil with a 10X loupe having a millimeter scale divided to 0.1mm. At 10X, the divisions are angularly so well separated that achieving 0.05mm is child's play.

While perhaps not quite significant enough to worry about in most cases, there is something one should be at least aware of. If the aperture-limiting stop is located relatively near to the eyepiece's field stop, its sharply-imaged edge will lie a small distance *behind* the image of the objective. This is because the eyepiece is a projection lens, and the closer to it an object lies, the farther back will the image formed by it lie.

You can verify this by simply looking into the exit pupil with a 7-10X loupe. Don't have one? Then an eyepiece of 20-30mm f.l. will do. You may have to turn it in reverse and/or remove its lower barrel, if focusing/viewing is difficult. As you slowly move the magnifier closer/farther, you'll see that you are successively bringing into focus various internal bits, such that inward you focus closer to the objective, and outward you focus closer to the prisms.

In the worst cases, where the offending stop lies *much* closer to the field stop than the objective, it's possible for the *new* exit pupil to lie far enough behind the image of the objective that the resulting slight increase in its magnification could become of some significance. But I'll have to do a more detailed analysis on this provisional theory!

At any rate, there *will* be a slight but artificial increase in the eye relief. My gut tells me that for a 20mm eyepiece this *could* amount to as much as a couple of millimeters. (While I highly doubt that this increased measure would be the reported value, I wouldn't discount the possibility entirely in some cases.)

#57 Tony Flanders

Tony Flanders

    Hubble

  • *****
  • Posts: 18,776
  • Joined: 18 May 2006

Posted 23 September 2009 - 06:43 PM


This would only allow you to measure exit pupil IF you took the measurement at precisely the eye relief distance, which adds another measure of difficulty. As you move even 6" to 12" away, the circular projection grows to several inches large.


I don't see that. If you focus the device for infinity and throw a collimated beam into the objective, you should get a collimated beam out the eyepiece.


Okay, tossing this over in the back of my mind at work, I saw a potential flaw in this reasoning -- which I verified by experiment when I got home.

What I said is technically true, of course. If you throw a collimated beam into the objective, you get a collimated beam out that eyepiece. That's the *definition* of infinity focus. But in real life, no beam of light can be perfectly collimated. What happens when the collimation is a little off?

When you shine a beam into the eyepiece of a telescope, the beam emerging from the objective has any miscollimation greatly reduced. But when you do it in the opposite direction, any error is greatly increased.

Let's try an example -- pretty close to what I just did on my kitchen table. Take a 10x50 monocular with an f/4 objective, focused at infinity. The objective's focal length is 200 mm, the eyepiece's focal length is 20 mm, and they're spaced 220 mm apart.

Now put the light source 3 meters from the objective. Instead of coming to focus at the proper spot, the beam comes to focus 1/(1/200 - 1/3000) ~= 214 mm from the objective. That's 6 mm from the eyepiece instead of the proper 20 mm, so the light beam coming out of the EP diverges strongly -- as though it were coming from a source 1(1/6 - 1/20) ~= 9 mm inside the instrument. Yow!

Now try this the other way. Shine the beam into the eyepiece from a distance of 15X the eyepiece's focal length, or 300 mm. The light comes to focus 1/(1/20-1/300) ~= 21.4 mm from the eyepiece, or 198.6 mm from the objective. So the beam coming out of the objective diverges as though it were coming from a source 1(1/198.6 - 1/200) = 28,000 mm away. For practical purposes, that's almost as good as infinty. No wonder Glenn's method is relatively insensitive to where you measure the beam!

In theory, you can eliminate the divergence by focusing on the light source -- at the cost of modifying the internal geometry of the instrument. But in practice, it's simply not possible to focus sufficiently accurately when shining the beam into the objective. I managed to reduce the size of the beam far from the EP quite a lot by careful focusing, but I never got it close to the size as measured at the exit pupil.

Moreover, even if you could focus perfectly, you'd still have to deal with the fact that the light source is not infinitessimal. But that's an entirely different kettle of fish.

#58 GlennLeDrew

GlennLeDrew

    Hubble

  • *****
  • topic starter
  • Posts: 15,850
  • Joined: 17 Jun 2008

Posted 23 September 2009 - 07:08 PM

Interesting analysis, Tony! And it nicely augments what I was cobbling up just now, probably while you were performing your experiment.

#59 zanti-misfit

zanti-misfit

    Explorer 1

  • -----
  • Posts: 88
  • Joined: 17 Aug 2009

Posted 24 September 2009 - 02:11 AM

Thanks for the info Glenn. I'm going to do some more testing tomorrow and refine my process a tad. :)

#60 milt

milt

    Viking 1

  • *****
  • Posts: 603
  • Joined: 13 Sep 2004

Posted 24 September 2009 - 08:22 PM

So the beam coming out of the objective diverges as though it were coming from a source 1(1/198.6 - 1/200) = 28,000 mm away.

Tony, you have hit on the elegance of Glenn's method. I will try to put it even simpler for anyone still reading:

The definition of system magnification is the ratio of the angle that an off-axis ray exits the eyepiece to the angle that it entered the objective.
Going from objective to eyepiece the angle gets multiplied by the magnification, while going from eyepiece to objective the angle gets divided by the magnification.

The uncertainty in the number of off-axis rays picked up by the objective and passed through the field stop was what concerned me about Ed's illumination test.

Milt

#61 Gordon Rayner

Gordon Rayner

    Gemini

  • *****
  • Posts: 3,303
  • Joined: 24 Mar 2007

Posted 25 September 2009 - 04:58 PM

"The uncertainty in the number of off-axis rays picked up by the objective and passed through the field stop was what concerned me about Ed's illumination test." Do you mean that the holographic target projecting laser does not have a uniformly illuminated pattern? Are you still "concerned"? Did I miss an applicable posting? The pattern is probably two lines, more or less interrupted, at right angles to each other, with a scattering of illuminated spots in the remainder of the circular pattern which is projected into the objective at an angle inclined to the optical axis, which is likely a compound angle ("skew"?) if the laser is handheld.

Is the concern about the blank portions of this projected, holographically generated, pattern? Would a solution be a widefield collimator, that is, a target collimator with a very large target, brilliantly and importantly, uniformly, illuminated,with an objective large enough to avoid vignetting within the collimator,and large enough to uniformly illuminate all parts of the binocular objective, shining into the objective of the binocular? The resulting target pattern, after leaving the eyepiece, would show any internal blockage, and the vignetting, on a larger scale than the flashlight test which works in the other direction.

Or, put the binocular on an astrophotography tracking mount, with a rotation support for the binocular to rotate around its yaw axis and use the(filtered?) projected solar image ( or defined part(s) of it(to avoid limb darkening)) to look for blockages and measure vignetting. One could also point the eyepiece at the sun and measure the effective aperture of the binocular with a ruler near the objective, ff the LeDrew test of this thread.

I just now did this reversed binocular solar test with a B&L MK21 WW II 7 x 50, freehand, and measure 50mm. A Swift Neptune Mark II 7 x 35, model 802, marked SK inside an equilateral triangle with truncated corners, gives 35mm. To my knowledge ,I do not have any dubious specimens likely to fail .

Another method, for quantitative vignetting measurement, might be to mount a small LED flashlight on an angularly adjustable base, shine it through the objective at defined and recorded angles, and use a photometer to record the throughput at the corresponding positions in the post-eyepiece pattern.

Has anyone dug out the NavOrd or BuOrd postwar report by Howard Coleman at Penn State? I could find my printed copies, but have not yet looked. I think that it may be online in some of the Abrahams material also. There maybe an applicable optical bench-like setup there.

#62 milt

milt

    Viking 1

  • *****
  • Posts: 603
  • Joined: 13 Sep 2004

Posted 25 September 2009 - 08:46 PM

Do you mean that the holographic target projecting laser does not have a uniformly illuminated pattern?....Did I miss an applicable posting?


Hi Gordon,

My concern was not for the laser test although I confess when did it I ignored the holographic portion of the image and only used the central beam. This is actually the only test I fully understand because it is a proxy for a single light ray entering one point on the objective at a defined angle.

The one that had me concerned was Ed's recent experiment using a non-coherent light source shining down a reflective tube into the objective and then measuring the total light output from the eyepiece. It seemed that all other things being equal, this would favor binoculars with larger field stops at the focal plane.

Another method, for quantitative vignetting measurement, might be to mount a small LED flashlight on an angularly adjustable base, shine it through the objective at defined and recorded angles, and use a photometer to record the throughput at the corresponding positions in the post-eyepiece pattern.



I don't believe an LED flashlight is a coherent light source either...? In any case I think there is a place for both Glenn's light-in-the-eyepiece test to quickly measure true aperture and Ed's laser-in-the-objective test to ray-trace the system.

Milt

#63 aa5te

aa5te

    Genial Procrastinator

  • *****
  • Posts: 1,918
  • Joined: 30 Aug 2008

Posted 26 September 2009 - 11:19 PM

I just did this test on a few of mine:

Tasco InFocus 10x50 = 39mm (ha ha! $10 K-mart clearance)
Pentax 20x60 PCF WP = 59mm
Apogee and Burgess Optical 25x100 = both around 95mm
Meade DS-2102AT 102mm refractor = 102mm (I did this one because the tube is only 90mm in diameter, but the lens cell is a few inches long (and wider), giving the light cone enough length to converge enough to not be clipped by the tube)

#64 GlennLeDrew

GlennLeDrew

    Hubble

  • *****
  • topic starter
  • Posts: 15,850
  • Joined: 17 Jun 2008

Posted 26 September 2009 - 11:28 PM

Regarding that Tasco InFocus 'so-called' 10X50... Is the effective aperture the same for both halves? What's the exit pupil diameter? (This would verify the test, and furthermore would indicate whether the magnification is indeed 10X.) But no matter how you look at it, 39mm is one heck of a small effective aperture for a 50mm!

#65 KennyJ

KennyJ

    The British Flash

  • *****
  • Posts: 37,648
  • Joined: 27 Apr 2003

Posted 27 September 2009 - 03:57 AM

Glenn ,

The so - called 10 x 50 Bresser Porros traditionally sold off in Christmas time sales at Lidl superstores here in the UK for £9.99 also only have a true aperture of around 40mm.

You may have seen the reference 10 x 50 ( 40 ) used a few times by various members here who own the model , which performs surprisingly well for astro use .

From my own non - technical tests when I had one ( I bought three to pass on to others ) I noticed the exit pupils were very close to 4mm in diameter and the magnification just a tad less than real 10x binoculars -- probably closer to 9.5x than 10x if expressed to the nearest 0.5x .

Indeed I once refered to the model here as 9.7 x 39 :-)

Kind Regards
Kenny



#66 aa5te

aa5te

    Genial Procrastinator

  • *****
  • Posts: 1,918
  • Joined: 30 Aug 2008

Posted 27 September 2009 - 10:09 AM

39mm on both halves is verified! The exit pupil is right at 4mm, so the magnification would be 9.75x - close enough to 10x for government work. I guess this partially explains why I noticed such a large jump in brightness when comparing these to the Pentax pair - instead of the expected 10mm gain, there is actually a 20mm gain (not to mention superior quality, etc.).

#67 Gordon Rayner

Gordon Rayner

    Gemini

  • *****
  • Posts: 3,303
  • Joined: 24 Mar 2007

Posted 27 September 2009 - 10:59 PM

I did the solar test on a Simmons "10x50" fixed focus Chinese "made in Hong Kong", says a label on the box. 39-40mm for both sides. I seem to recall that the Kamakuras had something to do with those. They were making many/all of the products sold by Bresser.


A Bushnell IF 8 x 40 wide field, probably from the 1950's, measures 40mm.
A Fuji Meibo 7 x 50 IF MTR ( from which I removed the deteriorated, sticky rubber covering) measured 50mm.

The solar test can be performed with two hands: one to hold the ruler, and one to hold the binocular. A piece of cardboard, plastic, wood, etc, with an opening through which the eyepieces can protrude toward the sun, will shade the ruler from direct sunlight, if that should be required.

#68 harbinjer

harbinjer

    Viking 1

  • *****
  • Posts: 510
  • Joined: 17 Dec 2008

Posted 28 September 2009 - 10:28 AM

It tried a Tasco Zip focus 10x50, and got about 44mm.
Also an old Yashica 7x50, which came up as 50mm.
The Audubon Raptor 8x42 seemed around 39mm.

#69 Rick

Rick

    Gemini

  • *****
  • Posts: 3,291
  • Joined: 12 Apr 2005

Posted 28 September 2009 - 10:24 PM

After reading about the Canon 10x42L IS, I tested my 18x50IS and both objectives are perfectly circular 50mm with both exit pupils <= 3mm. Close enough for me!

thx Glenn!
Rick

#70 richtea

richtea

    Ranger 4

  • -----
  • Posts: 309
  • Joined: 01 Feb 2005

Posted 29 September 2009 - 03:19 AM

Hi Glen and all

Great little test this even for non experts like myself
Tried on an old Swift 10 x 42 Ultralite porro and Bresser/Meade 10 x 50 budget porro last night

The Swifts were as close to 42mm as my eye could check and the Bresser/Meades were about 41mm as others have reported
Strange that so many beginners will purchase a binocular on manufacturers spec and/or advice on what spec is best when in fact the truth maybe quite different

Thanks
Regards
RichT

#71 BarrySimon615

BarrySimon615

    Pa Bear

  • *****
  • Posts: 4,495
  • Joined: 01 Mar 2004

Posted 19 October 2009 - 06:19 PM

I have tested some of my binoculars using the method discussed. I also had the advantage of having some disposable paper rulers used for measuring the length of newborns in the hospital. The rulers have both millimeter and inches. I simply taped a ruler to a door, set up a binocular on tripod and squared the binocular path to the door (leveled the optical axis of the binoculars and measured the distance of each side of the objective barrels to the door to make sure everything was aimed correctly.

Anyway here are my finding with perhaps the binocular that would be of most interest listed first -

Miyauchi BJ100 Galaxy 20x100 (and also measured with 26x eyepieces) - 98 mm, note the projected image of the objective was clean, sharp, evenly illuminated and did not show and chord cut off.

Celestron SkyMaster 25x100 - 92 mm, the projected image was edge fuzzy and also had the appearance of an eyeball with the eyelid partially down and covering the top of the image for both projected objectives. I played with this awhile thinking at first that it was my fault in how I was holding the light source, but I could not get this shading on the top half of both projected images to go away - hence very uneven illumination. When looking at the exit pupil, both looked round without any problems noted. Collimation is good, but the 92 mm aperture coupled with the very uneven illumination (eyelidding) is a concern. Maybe these binoculars would work for the Simpsons or Garfield the cat.

Pentax PCF Type V 16x60 - full 60 mm projected image, evenly illuminated, no issues

Nikon Action Extreme 10x50 - 49 mm, good projected image

Orion Savannah 10x50 - 49 mm, good projected image

Fujinon Series 2000 center focus 10x50 - 45 mm with chord cutoff

Carton Adlerblick 7x50 - 50 mm with clean projection and even illumination

Carton Adlerblick 8x42 - 42 mm with clean projection and even illumination

Oberwerk Mariner 8x40 - 32 mm

Fujinon Series 2000 7x35 - 33 mm with significant chord cutoff

I have a few more I could test but probably none of any real astronomical interest.

Barry Simon

#72 EdZ

EdZ

    Professor EdZ

  • *****
  • Posts: 18,849
  • Joined: 15 Feb 2002

Posted 20 October 2009 - 04:51 AM

My measurements agree with yours exactly or within 1mm on these

Celestron SkyMaster 25x100 - 91 mm
Pentax PCF Type V 16x60 - full 60 mm
Nikon Action Extreme 10x50 - 48 mm
Oberwerk Mariner 8x40 - 32 mm

edz

#73 Patrik Iver

Patrik Iver

    Apollo

  • -----
  • Posts: 1,038
  • Joined: 29 Jul 2003

Posted 23 October 2009 - 06:41 AM

Checked the effective aperture of the binoculars that happened to be in the house just now:

Pentax Papillio 6.5x21 reverse porro: ~20 mm, some minor prism cut-off.

Leupold Yosemite 6x30 porro : ~28 mm, perfectly round.

Canon 10x30 IS porro-2: ~29-30 mm, perfectly round.

Minolta Activa 7x35 porro: ~33 mm, some prism cut-off.

Telescope Service (Kunming) Triplet 8x42 roof (not phase coated): Left barrel ~37-38 mm with largeish prism cut-off. Right barrel ~38 mm with less prism cut-off.

Zeiss Victory FL 7x42 roof: ~41-42 mm, perfectly round.

Kronos 26x70 porro: 70 mm, or maybe even very slightly larger, perfectly round.

#74 RichD

RichD

    Vanguard

  • -----
  • Posts: 2,144
  • Joined: 08 Nov 2007

Posted 23 October 2009 - 07:13 AM

I see the uneven illumination on my Garrett 30x100 too Barry. Like you, I thought it was the way I was holding the light but it seems the problem is with the bino.

The eye analogy is a good one, and no matter how I present the light source to the eyepiece I can't get full illumination like I do with the Fuji 16x70.

Aperture however seems to be around 98mm, but I will recheck this given your result.

#75 Patrik Iver

Patrik Iver

    Apollo

  • -----
  • Posts: 1,038
  • Joined: 29 Jul 2003

Posted 23 October 2009 - 01:01 PM

Kronos 26x70 porro: 70 mm, or maybe even very slightly larger, perfectly round.


I just re-read this thread from the beginning, and noticed that KBK measured his Kronos 26x70 to be 70-71 mm, i.e. in accordance with my measurement.

I thought it a bit strange that they would appear to be slightly larger than specified, so I thought I would measure the objective lens diameter.

The trim rings around the ends of the barrels are on my sample between 68.86 and 68.95 mm (slighly out of round, measured with a digital caliper). I removed the trim rings and measured the objective retaining rings directly and found them to be between 69.98 and 70.16 mm (again slightly out of round).

The trim rings were in place when I measured the effective aperture using the flashlight method. Yet I see atleast 70 mm effective aperture with the trim ring forming a 69 mm aperture stop just a short distance from the objective lens. And KBK's observation seems to agree with mine.

I suppose the easiest explanation would be that I measured the effective aperture incorrectly (parallax or some such geometrical error)? I know the binos were set to infinity focus.

Could the fact that the light source is not an absolute point source cause a slightly too large reading which is more readily apparent in large aperture binoculars?

Any other ideas?


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics